The next generation of AI chatbots is personal. How to get it done and how it will affect the behaviour of your customers

Feb 27, 2026 | AI

By Laura Bjerre Schwalbe

}

Reading Time: 3 minutes

Back to blog

Chatbots are no longer experimental add-ons - for commerce teams, they should serve as a core part of your digital sales experience.

Recent industry reports show that nearly 70% of consumers have interacted with a chatbot in the past year, and adoption continues to accelerate across e-commerce, finance, travel, and subscription businesses. But the real change isn’t volume - it’s behaviour.
What started as simple FAQ automation is now evolving into something far more strategic: bots are increasingly acting as shopping assistants, not just support tools.

The shift is driven by two forces:

Customers’ behaviour is molded by the increased use of conversational experiences from LLMs such as OpenAI, Claude, Gemini, etc.
Brands have access to real-time behavioural data and AI-driven personalisation

LLMs can write and provide you with answers, but they don’t have long-term memory. They can’t reason deeply or learn on the fly. Everyone can use the same LLM models.

The winners will be those who connect the dots from raw intelligence to how intelligently it’s wired together.

The future is about conversational search

Connecting your chatbots to your own data gives them real power.

A Model Context Protocol (MCP) server acts as a real-time interface between your data sources and the chatbot. In short, it tells the chatbot what each shopper expects and likes - live and in real-time, to make the chatbot able to present relevant products — quickly, data-driven, and consistently across all channels.

A modern chatbot should sit inside your ecosystem of CRM systems, recommendation engines, and customer data platforms (CDP). When connected properly, it becomes a live interface to your customer intelligence — capable of guiding users, recommending products, resolving friction, and driving measurable revenue.

So the question is: how personal can your chatbot become?

Below are key areas where personalisation transforms a bot from a support cost into a growth engine / conversion driver.

Ways to improve your chatbot experience

Product and content recommendations
The most immediate way to personalise a chatbot is through intelligent recommendations.
With an MCP layer, a chatbot will be able to look for the most relevant products at lightning speed. Gone are the days of navigating through complex menus, product list pages, and the like. The chatbot will take care of that for you.
CDP + behavioural data: From session context to customer memory
The chatbot reacts to more than the current message — it responds based on memory. Now, the search experience starts becoming a conversation. The chatbot will ask follow-up questions, tailored to your customers’ ID. With a CDP, the search will have far more memory about the users, so every session feels smarter.
Steering the AI without losing relevance
Personalisation doesn’t mean losing commercial control. With a merchandising layer in the MCP system, you decide which products the chatbot prioritises — for example, seasonal items or strategic campaigns, while still keeping recommendations relevant to the individual customer.
After-sales, service & expansion
Personalisation beyond the purchase: A personalised chatbot becomes a lifecycle channel that supports retention, loyalty, and expansion. Use the memory from chats to inform customer service and sales about what users are searching for to eliminate gaps between sales channels.

We know. It sounds techy, nerdy, and a little bit complicated. But it’s simple, as long as you follow these steps and work with it as an integral part of your strategy.

USE CASE

Thibaut searches for Red Wine on La Boutique du Vin’s webshop.

The chatbot knows:

Thibaut bought Bourgogne the previous 3 times on the site. He has also purchased Californian, Italian and other French wines last year.
He buys mostly white wines from March-September and red wines from November-January.
His price range is 300-600DKK, mostly in the 300-400DKK category.

Thibaut searches for Red Wine in the chatbot.

The chatbot asks:

"Looking for another Bourgogne Pinot Noir around 350 DKK, or want to try something similar from another region?"

It shows 3 relevant bottles immediately.

The chatbot combines Thibaut’s preferences with his historical purchase data and seasonal buying behaviour.

Result:

Less browsing. Faster purchase. Higher conversion.

💡 Explainer:

CDP - A Customer Data Platform brings all customer data into one place to understand customers better. It helps businesses deliver personalised shopping experiences using their own data.

LLM - A Large Language Model is AI that understands and generates human-like text for customer interactions. In commerce, it enables support, recommendations, and personalised shopping.

MCP - A Model Context Protocol connects AI to business data and systems in real time. In commerce, it lets AI use product, customer, and order data to automate experiences.

How to get it done:

Depending on where you are on your chatbot maturity journey, here are five ideas on how to improve. One step at a time.

Launch a chatbot.
Within weeks, we can set up a chat widget connected to your product catalog, including features such as AI overview, intent understanding, relevance ranking, and speed optimisation.
Implementation of memory
Integrate your CDP so the chatbot knows purchase history, preferences, and behaviour. Now recommendations become personal, not generic.
Recommendations
Combine stock levels, margins, and campaigns with customer preferences. This turns the chatbot into a sales assistant that supports your merchandising strategy.
Work (pro)actively
See what your users search for, how they convert, and where AI needs tuning. Add more channels, smarter segmentation, and calculation of CLV.
Use chatbot insights across channels
Syndicate your data with platforms such as Meta, Google, etc. to create smarter campaigns and a significant boost in marketing ROI.

This blog post was written in collaboration with our partners at Alpha Solutions, dedicated experts in DXPs, CDPs, CMS, and all things digital experience.